Fed up with Dreamforce news yet? If the answer is “yes”, I don’t blame you. The noise over the past week from Dreamforce has been overwhelming, and with so much content, it’s hard to know what matters or what’s connected.
I wanted to take a minute to collate the biggest AI takeaways from Dreamforce in a single article, in the hope that it will give you a consolidated view of the announcements specific to AI, and how they piece together.
1. Agentforce
The big announcement at Dreamforce was Agentforce. Agentforce delivers autonomous agents to Salesforce, capable of operating independently without needing external control or continuous intervention. An Agent can make decisions based on predefined rules or learning algorithms – much smarter than a chatbot, let’s put it that way!
Copilot Builder has basically been rebranded to Agent Builder which is where you configure your agents.
You’ll get different types of agents out of the box, including those designed for sales and service.
2. Atlas Reasoning Engine
Atlas is the reasoning and learning engine behind Agentforce, otherwise known as the “chain of thought”. Atlas keeps “looping” until it’s confident it can achieve the goal that’s been set by your organization in line with the user’s request, whether that user is someone internal in your workforce or a customer/prospect. It then performs “reinforcement learning” based on human feedback – in other words, with every interaction, Atlas will get smarter for your own organization.
From a security standpoint, Altas respects the sharing model as defined at the application layer (e.g. which of these sales reps have access to which customer records?). As explained by one of the leaders close to this processing engine, Atlas is Salesforce’s upgraded copilot engine.
The diagram below is helpful in understanding the core components of the process – the ultimate aim being to generate precise and factual results.
Version 2 of Atlas is already in testing with a small number of customers, so it’s pretty safe to say that Atlas is only going to get better.
3. Data Cloud Underpins Your AI Success
Data Cloud is essential to use Agentforce, underpinning Atlas (the reasoning engine). So, my question is, what if you don’t have Data Cloud? As excited as I am by the AI innovations, all I can hear is the “ka-ching” of a cash register in my head when I think of all the products I need to use these features…
If you want to take advantage of what Agentforce has to offer, you’re going to need Data Cloud. One way to do this is to get Salesforce Foundations, a free* feature set that brings together new sales, service, marketing, commerce, and Data Cloud features, and also includes an updated user interface. Salesforce’s goal for Foundations is to give organizations a flavor of how powerful AI can become when using all these products together and to get Agentforce into the hands of customers.
*Free for all Sales and Service Cloud users on enterprise licenses or above. Note that some features have limits imposed on them.
4. AI Must be for Everyone
If AI is the next revolution in the way we work, we must all adapt. This means learning in-demand AI skills, not just to get jobs, but also to make sure we don’t break things along the way. The lack of AI knowledge terrifies me and I hear horror stories all the time about people putting intellectual property or personally identifiable information into public-facing GPTs like Chat GPT.
To support the current and upcoming wave of AI tools and implementations, Salesforce has announced that they’re investing $50 million into initiatives designed to upskill workforces in a bid to address the AI skills gap and democratize AI learning.
You can take the AI Associate and AI Specialist certifications for free until the end of 2025.
Summary
Those were my four key AI takeaways from Dreamforce ‘24. There was plenty of other news and there’s a lot more information on each of these topics, but at a high level, this is what you need to know!